Article
pubs.acs.org/Langmuir
Complementary Surface Second Harmonic Generation and Molecular
Dynamics Investigation of the Orientation of Organic Dyes at a
Liquid/Liquid Interface
Denis Svechkarev,†,‡ Dmitry Kolodezny,‡ Sandra Mosquera-Vázquez,† and Eric Vauthey*,†
†
Department of Physical Chemistry, University of Geneva, 30 quai Ernest-Ansermet, 1211 Geneva, Switzerland
Institute for Chemistry, V. N. Karazin Kharkov National University, 4 Svobody Square, 61022 Kharkov, Ukraine
‡
S Supporting Information
*
ABSTRACT: The second-order nonlinear response of two dyes
adsorbed at the dodecane/water interface was investigated by
surface second harmonic generation (SSHG). These dyes consist of
the same chromophoric unit, 2-pyridinyl-5-phenyloxazole, with an
alkyl chain located at the two opposite ends. The analysis of the
polarization dependence of the SSHG intensity as usually
performed points to similar tilt angles of the two dyes with respect
to the interface but does not give information on the absolute
direction. Molecular dynamics (MD) simulations reveal that both
dyes lie almost flat at the interface but have opposite orientations. A refined SSHG data analysis with the width of the
orientational distribution yields tilt angles that are in very satisfactory agreement with the MD simulations.
1. INTRODUCTION
Interfaces between two immiscible liquids are ubiquitous in
nature and play a crucial role in many technological
processes.1−3 Their specific properties, which often diverge
from those of the constituting phases, arise from the asymmetry
of the forces acting on this frontier region and from the
resulting anisotropic orientations of the interfacial molecules.
Therefore, molecules adsorbed at a liquid/liquid interface
experience microenvironmental characteristics, such as friction
or local electric field, that differ from those of the bulk phases,
which are usually described by macroscopic quantities such as
viscosity or dielectric constant. This in turn can lead to
chemical reactivity that deviates from that observed in bulk
solutions. For example, several chemical reactions involving
water-insoluble compounds have been shown to proceed faster
at the interface with water than in pure organic solvents.4−6
Knowledge of the orientation of the molecules adsorbed at
the interface is of primary importance for rationalizing their
chemical reactivity. However, this information is not readily
accessible. In order to probe only the adsorbed solute
molecules and not those dissolved in the bulk phases,
experimental techniques with high interfacial selectivity have
to be used. This is the case of nonlinear optical techniques such
as surface second harmonic generation (SSHG)7−10 and surface
sum frequency generation (SSFG),11,12 which probe the
second-order nonlinear optical susceptibility, χ⃡(2). Within the
dipolar approximation, χ⃡(2) vanishes in centrosymmetric media
but not at the interface between two isotropic liquids.
Moreover, χ⃡(2) is frequency-dependent and exhibits resonances
at frequencies corresponding to one- or two-photon transitions.
SSFG, which is mostly used with one of the optical fields in the
© 2014 American Chemical Society
infrared to study vibrational resonances, has proven to be
particularly powerful for investigating naked interfaces (i.e.,
interfaces without a solute adsorbed).13,14 On the other hand,
SSHG is usually used in the visible region in resonance with
electronic transitions of the adsorbed molecules.15−19 Electronic SSFG has recently been demonstrated to be a promising
alternative for studying adsorbates at air/liquid interfaces.20,21
Information on the orientation of the adsorbate can be deduced
from polarized SSHG experiments.22−25 However, a number of
approximations normally have to be made when analyzing the
resulting data to extract orientational information. One of them
is the use of a Dirac δ function to describe the angular
distribution of the molecules at the interface, which obviously
loses its physical meaning when going from a theoretical
approach to real systems. This issue has been addressed in
several studies,23,25−28 but the use of an angular distribution is
still not systematic. Recently, a novel approach was suggested to
obtain the orientational distribution of the transition dipole
moments of the adsorbate using a ratio of second harmonic
intensities recorded before and after excitation of the molecule
with a circularly polarized pump pulse and using particular sets
of polarization for the probe and signal fields.29 However, when
pump−probe studies are not feasible, it is not possible to obtain
this information from SSHG polarization profiles alone.
Additionally, as SSHG is inherently a homodyne technique,
the signal intensity is proportional to the square modulus of the
susceptibility, |χ⃡(2)|2. As a consequence, no information on the
Received: August 5, 2014
Revised: October 30, 2014
Published: October 31, 2014
13869
dx.doi.org/10.1021/la503121g | Langmuir 2014, 30, 13869−13876
Langmuir
Article
of a Czerny−Turner spectrograph (Shamrock 163, Andor) equipped
with a cooled CCD camera (Newton 920, Andor). The illuminated
pixels were vertically binned and summed over the wavelength range
of interest. In all of the experiments, the polarization of the probe
pulse was controlled with a half-wave plate, and the p, s or −45°
polarized component of the SSHG signal was detected. The
nonresonant contribution to the SSHG signal measured without dye
in the aqueous phase was found to be negligibly small.
Quantum-Chemistry Calculations. The ground-state geometries
and charge distributions of dodecane, N1, and N8 were optimized
using density functional theory (DFT) with the B3LYP functional and
the cc-pVDZ basis set as implemented in Gaussian 09.43 The transition
dipole moments of both N1 and N8 were calculated using timedependent DFT (TD-DFT) with the same combination of functional
and basis set. The hyperpolarizability tensor elements were computed
at the B3LYP/6-31G++(d,p) level at 800 and 1000 nm. These
wavelengths were chosen because the experimental and calculated S1
← S0 absorption maxima of both dyes are at 400 and 500 nm,
respectively, and thus correspond to two-photon resonances.
Molecular Dynamics Simulations. The MD model of the
dodecane/water interface was based on the all-atom CHARMM36
force field for the GROMACS package, version 4.6.5,44 with the threesite TIP3P water model.45 The bond lengths and angles for dodecane,
N1, and N8 were adapted to the CHARMM force field format with
PRODRG.46 The partial charges necessary to account for Coulomb
interactions were derived from the B3LYP/cc-pVDZ electron densities
by fitting the electrostatic potential to point (ESP) charges. All of the
covalent bonds were constrained to their equilibrium values using the
P-LINCS algorithm.47 The repulsion and dispersion terms of
nonbonded interactions were computed using the Lennard-Jones
potential energy function. Electrostatics was treated with a particlemesh Ewald model48 using a short-range cutoff of 1.0 nm and van der
Waals interactions switched off between 0.8 and 1.0 nm. The
temperature was maintained using the thermostat by Bussi et al.49
Periodic boundary conditions were applied, as well as the Berendsen
barostat.50 The simulations were performed with constant number of
particles, pressure, and temperature (NPT ensemble) using an
integration time step of 1 fs with the neighbor list updated every 20
fs. Chloride ions, available in the force field, were used as counterions
instead of tosylate ions.
First, a cell containing 267 dodecane molecules was created and
equilibrated during 1 ns. In a second stage, two 5 nm × 5 nm × 4 nm
cells were added and filled with water (∼6400 molecules) to obtain a 5
nm × 5 nm × 12 nm sandwichlike cell with the dodecane phase in the
middle and a water phase on each side along the z axis. Because of the
periodic boundary conditions adopted here, using only two layers
would have led to the formation of two new unwanted dodecane/
water interfaces. Additionally, the sandwichlike cell allowed
information about the dynamics of N1 and N8 to be gathered
simultaneously. The volume of the cell also allowed some simulations
to be started with the dye molecules located in the bulk phase. A 20 ns
trajectory was recorded in every case, allowing enough time for the
accumulation of data after adsorption of the dye molecules. Because
the motions of the atoms in the MD simulations are random, it was
necessary to prepare several starting geometries to analyze all of the
most probable cases and to obtain reliable statistical information.
Thus, in two cases, the molecules were standing vertically with either
their octyl or methyl tail pointing toward the dodecane phase, and
three more runs were made with the molecules initially lying flat with
respect to the interface. Trajectories were visualized with the VMD
package to analyze the behavior of the adsorbed dyes.51
Because of the periodic boundary conditions, the molecule could go
beyond the cell at one (or more) edges of the cell and virtually appear
at its opposite side. In terms of atomic coordinates, this means that the
difference between the x and y coordinates becomes bigger than the
cell size, thus artificially increasing the population of molecules lying
almost flat with respect to the interface. To address this issue, the
analysis algorithm was amended to detect abnormally large x and y
vector coordinates that exceeded a certain threshold, set to be half of
the molecule’s length. The dimensions of the cell at each step were
absolute orientation of the adsorbate (i.e., on which particular
end of the molecule is pointing up or down) can be obtained
unless heterodyne detection is performed.21,30
Alternatively, modern computational methods such as
molecular dynamics (MD) can be particularly helpful to gain
insight into the interfacial behavior of a solute molecule. Over
the past years, MD has been increasingly used to simulate
interfaces and interfacial processes,31−35 for example, adsorption and solvation,34,36,37 extraction,38 or the microenvironment
of adsorbates.39,40
To the best of our knowledge, the determination of the
orientation of a solute molecule at a liquid/liquid interface
using the combination of polarized SSHG experiments and MD
simulations has not been described to date. Here we report on
the application of such a combination to study the orientation
of two organic dyes consisting of the same pyridine−oxazole−
phenyl (POP) chromophoric unit but having an octyl tail
attached to the opposite ends at the dodecane/water interface
(Figure 1). The chromophore bears donor (alkyloxy) and
Figure 1. Structures of the dyes, key atom numbers, and calculated S1
← S0 transition dipole moments (green arrows, not to scale).
acceptor (quaternary nitrogen) moieties, thus giving a good
perspective for a substantial second-order nonlinear response.
The influence of the alkyl chain at both sides of the
chromophore on the adsorption at the interface as well as on
its orientation are discussed. The results of the MD simulations
are compared to the information extracted from the analysis of
SSHG polarization profiles and applied to go beyond the
assumptions usually used in such SSHG data analysis.
2. EXPERIMENTAL SECTION
Sample Preparation. The organic dyes, 1-methyl-4-(5-(4(octyloxy)phenyl)oxazol-2-yl)pyridinium tosylate (N1) and 4-(5-(4methoxyphenyl)oxazol-2-yl)-1-octylpyridinium tosylate (N8), were
synthesized in the group of Prof. A. O. Doroshenko (V. N. Karazin
Kharkov National University) and were used as received. Distilled
water and dodecane (pure from Acros Organics) were used without
special purification. The absorption spectra were recorded on a Cary
50 spectrophotometer (Varian) in 1 cm quartz cells. The samples for
the SSHG experiments were prepared by pouring 10 mL of distilled
water into a 4 cm × 4 cm × 4 cm quartz cell along with 100 μL of the
dye solution in distilled water (approx. 10−4 M) and then adding 12
mL of dodecane. All of the measurements were performed at room
temperature.
Surface Second Harmonic Generation. The SSHG measurements were performed using the experimental setup described
previously.41,42 In brief, the probe pulses were generated with a
noncollinear optical parametric amplifier (TOPAS-C, Light Conversion) pumped by a 1 kHz Ti:sapphire amplified system (Solstice,
Spectra-Physics). About 0.3 μW of the TOPAS output beam was
focused through the dodecane phase onto the interface at an angle of
incidence of about 70°, where it underwent total internal reflection.
The SSHG signal was then collected and focused onto the entrance slit
13870
dx.doi.org/10.1021/la503121g | Langmuir 2014, 30, 13869−13876
Langmuir
Article
derivatives, has π−π* character and corresponds to the S1 ← S0
electronic transition.52 Figure 2A also shows the stationary
SSHG spectra measured with N1 and N8 at the dodecane/
water interface by tuning the probe wavelength between 800
and 1000 nm. The band maximum is at around 440 nm for
both dyes, but the band of N1 is broader with a shoulder at
around 480 nm. Compared with the electronic absorption
spectra, both SSHG spectra are red-shifted by about 40 nm
(2200 cm−1) and are substantially narrower. Such shifts and
bandwidth differences have been observed previously with
other dyes such as coumarins.53−55 Whereas the shift has been
ascribed to a solvatochromic effect, the origin of the narrowing
has not been totally clarified. In the present case, the absorption
spectra of N1 and N8 measured in toluene are essentially the
same as those in water, and thus, environmental effects cannot
be invoked here. However, it should be noted that the
enhancement of the SSHG intensity in this spectral region
arises from a resonance with a two-photon transition, and
therefore, the SSHG spectrum should not only be compared to
the linear absorption spectrum but also to the two-photon
absorption spectrum, which has been shown to often differ
from the one-photon absorption spectrum.56 The SSHG signal
obtained without the dye in the aqueous phase was negligibly
small, indicating that the SSHG signal with the dyes has a
purely resonant character of dipolar origin.57 Therefore, the
signal intensity depends on the square modulus of the relevant
elements of χ⃡(2) and hence on the square of the number of dye
molecules contributing to the signal.19
Polarized SSHG Experiments. The dependences of the
intensities of the s, p, and −45° components of the SSHG
signal on the polarization of the probe field at 850 nm
measured with both dyes at the dodecane/water interface are
shown in Figure 2B and C. The polarization profiles measured
at 800 and 1000 nm had the same shape, within the limit of
error, excluding the contribution to the measured signal by
other species, such as aggregates.58 If the SSHG spectra in
Figure 2A were due to several species, the polarization profiles
should vary with the probe wavelength unless all of the species
have very similar χ⃡(2) tensors and orientations at the interface,
which is very improbable. The solid lines in Figure 2B and C
are the best fits to the data of the following equation describing
the SSHG intensity:23,59
then additionally recorded, and the translation procedure was applied
when necessary to bring the molecule back to the neighbor virtual cell.
To analyze the molecular orientation at the interface at each time
step along the trajectory, vectors going between atoms 10 to 31 and 30
to 9 (Figure 1) were chosen as the z axes for N1 and N8, respectively.
These vectors are close to the actual transition dipole moments.
Additionally, the xy molecular planes of N1 and N8 were defined using
atoms 10, 25, and 31 and atoms 30, 19, and 9, respectively. These
atoms were chosen to ensure that small torsional rotations of the rings
would have very little impact on the directions of the chosen vectors.
3. RESULTS AND DISCUSSION
Spectral Properties. The electronic absorption spectra of
the two dyes in water exhibit a broad band with a maximum at
approximately 400 nm (Figure 2A), which, as for its wellknown analogue 2,5-diphenyloxazole (PPO) and its numerous
(2)
(2)
(2)
ISSHG = C·|a1χXXZ
sin 2γ sin Γ + (a 2χXXZ
+ a3χZXX
(2)
(2)
+ a4χZZZ
) cos2 γ cos Γ + a5χZXX
sin 2 γ cos Γ|2 Ip2
(1)
where C is a constant that depends on the solvents and the
probe wavelength; ai (i = 1−5) are the Fresnel coefficients;
(2)
(2)
χ(2)
XXZ, χZXX, and χZZZ are the three independent nonzero
elements of the second-order susceptibility tensor for a material
with a C∞v symmetry, as is the case for an interface,8 with the
subscripts representing the Cartesian coordinates in the
laboratory frame (Figure 3); γ is the polarization angle of the
probe field (γ = 0° and 90° for p and s polarization,
respectively); Γ is the angle of the polarization component of
the SSHG signal (Γ = −45°, 90°, and 0° for the −45°, s, and p
components, respectively); and Ip is the intensity of the probe
pulse. The Fresnel coefficients were first calculated as described
in the Supporting Information, and the relative magnitudes of
the three tensor elements were adjusted by using all three
polarization profiles simultaneously. The resulting values are
listed in Table 1. These values point to substantial differences
Figure 2. (A) Electronic absorption spectra in water and stationary
SSHG spectra (p component) measured at the dodecane/water
interface (the polarization of the probe pulse was at 45°). (B, C)
Dependences of the intensities of the s, p, and −45° components of
the SSHG signals measured at 425 nm for (B) N1 and (C) N8 on the
polarization angle of the probe field (γ) at the dodecane/water
interface. The solid lines show the best fits of the data to eq 1.
13871
dx.doi.org/10.1021/la503121g | Langmuir 2014, 30, 13869−13876
Langmuir
Article
(2)
χZZZ
=
N
⟨cos3 θ ⟩βzzz
ε0
(3b)
where the tilt angle θ and the other Euler angles, which describe
the orientation of the dye with respect to the interface, are
defined in Figure 3.
The orientation parameter D, which directly depends on the
tilt angle θ, can be calculated from the following combination of
the χ⃡(2) tensor elements:53,59
(2)
D=
between the interfaces with N1 and N8 adsorbed, pointing to
distinct orientations at the interface. However, as χ⃡(2) is a
macroscopic quantity, it does not give direct access to this
orientation. For this, one has to consider the microscopic
equivalent, which is the second-order polarizability or hyperpolarizability tensor, β⃡:
(2)
where N is the interfacial adsorbate density and the angle
brackets indicate an average over the molecular orientations.
Relating χ⃡(2), represented in the laboratory frame, to β⃡ first
requires a molecular coordinate system to be defined. As both
N1 and N8 belong to the C1 point group, the long molecular
axis of each dye, defined as discussed above, was chosen as the z
axis. This axis is also close to parallel to the S1 ← S0 transition
dipole moment (Figure 1). It has previously been shown that
the conformation of a polar tail on a chromophore can
noticeably influence the direction and size of the transition
dipole moment.60 A similar effect due to the flexibility of the
chromophore could not be ruled out here but was assumed to
be negligible. The conjugated π-electronic system thus is
located in the xz plane (Figure 3). As the strong S1 ← S0
transition in the POP chromophore has π−π* character, almost
no electronic density redistribution occurs along the y axis upon
excitation. As a consequence, the hyperpolarizability tensor
elements with a y index should make a negligible contribution.59 Quantum-chemical calculations of β⃡ at the resonance
wavelength were carried out to determine whether the number
of relevant tensor elements could be further reduced. The
results listed in Table S2 in the Supporting Information indicate
that only one tensor element, βzzz, is significant, in good
agreement with predictions based on symmetry considerations.
In this case, eq 2 can be developed as59
(2)
χZXX
=
N
⟨sin 2 θ cos θ ⟩βzzz
2ε0
(4)
With the values of the χ⃡(2) tensor elements obtained from the
SSHG polarization profiles, D amounts to 0.39 ± 0.03 and 0.21
± 0.04 for N1 and N8, respectively. The determination of the
angle θ from D requires knowledge of f(θ), the distribution
function of θ. In most previous works, a Dirac δ function was
used, leading to the simplest case of D = cos2 θ. Making this
assumption here results in θ values of 51.8 ± 1.8° (or 128.2 ±
1.8°) and 63.1 ± 2.6° (or 116.9 ± 2.6°) for N1 and N8,
respectively. These two θ values for each dye reflect the two
possible directions of the transition dipole moments, i.e.,
toward the dodecane phase or the aqueous phase. As discussed
above, the absolute orientation of the adsorbate cannot be
deduced from the polarized SSHG data. For the sake of
simplicity, we will henceforth only consider the smallest θ
value, which describes the tilt angle of the long molecular axis.
The absolute orientation of the molecule will be mentioned
when necessary. Such a δ distribution function as assumed here
is highly improbable for dyes adsorbed at a liquid/liquid
interface, and therefore, the tilt angles obtained with this
assumption correspond to lower-limit values.23 Further
information about the orientations of N1 and N8 was obtained
by performing MD simulations.
MD Simulations. Some of the simulations were performed
with the dyes initially located in the aqueous bulk phase. After
approximately 1−4 ns, both N1 and N8 were adsorbed at the
interface and stayed there for the remaining duration of the
simulation. The time at which the dye was adsorbed at the
interface was chosen as the starting point for the trajectory
analysis. In all of the simulations, the dyes had a definite
orientation relative to the interface, i.e., with the octyl chain
pointing up into the dodecane, down into the aqueous phase,
or along the interface. Figure 4A illustrates the variation of the
angle θ during 20 ns of simulation. Equilibration can be clearly
noticed after about 3 ns. All of the MD results shown in the
figures were obtained by starting the simulation with the octyl
chain up in the dodecane sulphas. The angle distributions
obtained from such trajectories could be well-reproduced using
a Gaussian function (Figure 4B,C):23
Figure 3. Laboratory and molecular coordinate systems and Euler
angles.
χ ⃡ (2) ∝ N ⟨β ⃡ ⟩
χ
⟨cos3 θ ⟩
= (2) ZZZ (2)
⟨cos θ ⟩
χZZZ + 2χZXX
⎡ (θ − θ )2 ⎤
0
⎥
f (θ ) = A exp⎢ −
⎣ 2(Δθ )2 ⎦
(3a)
(5)
Table 1. Relative Values of the Nonzero Tensor Elements of χ⃡(2) Obtained from the Fits of Equation 1 to the Polarization
Profiles
dye
χ(2)
XXZ
χ(2)
ZXX
χ(2)
ZZZ
N1
N8
(0.90 ± 0.02) + (0.99 ± 0.02)i
(1.57 ± 0.02) + (1.72 ± 0.02)i
(0.65 ± 0.02) + (0.81 ± 0.02)i
(1.36 ± 0.02) + (1.74 ± 0.02)i
(1.00 ± 0.02) + (1.00 ± 0.02)i
(1.00 ± 0.03) + (1.00 ± 0.03)i
13872
dx.doi.org/10.1021/la503121g | Langmuir 2014, 30, 13869−13876
Langmuir
Article
The Gaussian distribution f(θ) obtained from the simulations
can now be used instead of the δ function to extract the most
probable angle θ0 from the orientation parameter D (eq 4).23
The D value of 0.39 ± 0.03 obtained for N1 results now in θ0 =
61 ± 2°, whereas the D value of 0.21 ± 0.04 found for N8
yields θ0 = 76 ± 4° (see the Supporting Information for details
of the procedure). These values obtained by using the
physically more realistic angular distributions are now in better
agreement with the MD simulations (Table 2). The remaining
∼10° difference could be due to various factors: (i) the
direction of the S1 ← S0 transition dipole moment of the
adsorbed dye may differ from that calculated for the
equilibrium geometry; (ii) the local temperature in the
interfacial region irradiated by the probe beam might be higher
than assumed in the MD simulations, leading to a broadening
of the angular distribution; (iii) some specific properties of
interfacial water may not be well-reproduced by the MD
simulations; and (iv) interactions between nearby adsorbates
might influence their orientation.
As the hyperpolarizability tensors of both molecules are
dominated by a single element, βzzz, no information on the
Euler angle ψ, which represents the rotation around the main
molecular axis (Figure 3), can be obtained from the SSHG
polarization profiles. Figure 6A shows the variation of ψ for
both N1 and N8 during a 17 ns equilibrated MD trajectory.
Although the distribution of ψ is quite broad, there is a clear
domination of this angle around 90°, which corresponds to the
molecular xz plane facing the interface. One can also notice that
the distribution is broader for N8 than for N1 and that its
center, ψ0, is not the same in all trajectories (Table 2). This
difference in Δψ might arise from the ability of the methoxy
end of N8 to be in either the aqueous phase or the apolar
phase, the latter case requiring ψ to depart from 90°.
Additionally, the higher friction experienced by N1 as it
penetrates deeper into the dodecane phase, which is
substantially more viscous than water (ηdodecane = 1.37 cP vs
ηwater = 0.89 cP), might decrease the amplitude of the
fluctuations of ψ around its most stable value.
Figure 7 gives an even wider look at the dependence of f(ψ)
on the tilt angle θ. For molecules with small θ, i.e., with the
molecular axis almost perpendicular to the interface, f(ψ) is
almost random. However, as θ approaches 90°, the distribution
becomes increasingly localized around 90°, i.e., the dye tends to
have its molecular plane parallel to the interface. As a
consequence, the overall f(ψ) distributions shown in Figure
6B,C are due to the predominance of adsorbates with a large tilt
angle.
Figure 4. (A) Variation of the tilt angle θ along a whole MD trajectory.
(B, C) f(θ) distributions after equilibration with best fits to eq 5. The
full width at half-maximum (FWHM = 2(2 ln 2)1/2Δθ) is also given
for comparison.
where A is a normalization constant and Δθ and θ0 are the
width (standard deviation) and center of the distribution,
respectively. The Δθ and θ0 values obtained from trajectories
with different starting orientations are listed in Table 2. The
f(θ) distributions are essentially independent of the initial
orientation and indicate that the orientation with the octyl
chain pointing up into the dodecane phase is the most
probable, as expected from basic solubility considerations. This
implies that the POP chromophores of N1 and N8 have
opposite orientations at the interface.
Figure 5A,B shows MD snapshots of N1 and N8 adsorbed at
the interface. The MD simulations, in agreement with the
experimental data, indicate that N8 tends to lie more parallel to
the interfacial plane than N1. The pyridinium nitrogen, which
bears a partial effective positive charge, limits the penetration of
the dyes into the dodecane phase. Despite this, a major part of
N1 can be located in this nonpolar environment (Figure 5A).
For N8, on the other hand, only the octyl chain can penetrate
into the dodecane (Figure 5B). However, because of the
flexibility of this chain, the chromophic part of N8 is not fully
immersed in the aqueous phase but can adopt an orientation
almost parallel to the interface that probably reduces the
hydrophobic interactions and also allows the lipophilic methoxy
group to interact with the apolar phase (Figure 5C).
Table 2. Orientational Parameters Obtained from MD Simulations with Different Initial Orientations of the Dyes and from
Polarized SSHG Measurements
tilt angle
rotation angle
N1
initial orientation
trajectory length (ns)
octyl up
octyl down
horizontal
17
15.5
16.5
19
19
−
SSHG
a
a
N8
N1
N8
θ0
Δθ
θ0
Δθ
ψ0
Δψ
ψ0
Δψ
74
75
75
74
76
52
61
13
12
12
12
12
0
12
85
85
85
85
85
63
76
11
11
11
11
11
0
11
95
94
95
95
95
28
28
29
28
26
89
91
87
89
90
32
32
32
32
31
After equilibration at the interface.
13873
dx.doi.org/10.1021/la503121g | Langmuir 2014, 30, 13869−13876
Langmuir
Article
Figure 5. MD snapshots of typical configurations of (A) N1 and (B, C) N8 adsorbed at the interface. The phases are turned upside down for better
clarity of representation.
■
CONCLUDING REMARKS
The above investigation demonstrates that the combination of
SSHG spectroscopy and MD simulations is a very powerful
method for determining the orientation of organic molecules
adsorbed at liquid/liquid interfaces. We investigated two
molecules consisting of the same chromophoric unit but with
an aliphatic chain located at the two opposite ends. From the
analysis of the polarized SSHG measurements alone, only
relatively minor differences in the tilt angle between the long
molecular axis of the dyes and the interface normal were
obtained, pointing a priori to similar orientations. On the other
hand, MD simulations showed that the two dyes are oriented in
opposite directions, as expected from the positions of the alkyl
chain, and they allowed the distribution of the tilt angle to be
determined. Using the width of this distribution to refine the
analysis of the polarized SSHG measurements yielded tilt
angles that are in very satisfactory agreement with those
obtained from the MD simulations, although smaller in both
cases by about 10°. This difference is most likely due to the
various assumptions made in the SSHG data analysis and to
several effects not taken into account in the MD simulations.
Figure 6. (A) Variation of the angle ψ along the equilibrated part of a
trajectory. (B, C) Distributions with the best Gaussian fits.
■
ASSOCIATED CONTENT
S Supporting Information
*
Calculation of the Fresnel coefficients and details of the analysis
of polarized SSHG data; dependence of the distribution of ψ on
the tilt angle θ for N8; and topologies of dodecane and both
dyes used for MD simulations. This material is available free of
charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: [email protected].
Notes
The authors declare no competing financial interest.
■
Figure 7. Dependence of the distribution of the angle ψ on the tilt
angle θ for N1 obtained from MD simulations (statistics averaged for
all 87 ns of equilibrated trajectories available).
ACKNOWLEDGMENTS
We are thankful to Prof. P.-F. Brevet (Institut Lumière Matière,
Université Lyon 1) for fruitful discussions and to O. Borodin
and R. Iliashenko (V. N. Karazin Kharkov National University)
13874
dx.doi.org/10.1021/la503121g | Langmuir 2014, 30, 13869−13876
Langmuir
Article
(21) Yamaguchi, S.; Tahara, T. Heterodyne-Detected Electronic Sum
Frequency Generation: “Up” versus “Down” Alignment of Interfacial
Molecules. J. Chem. Phys. 2008, 129, No. 101102.
(22) Grubb, S. G.; Kim, M. W.; Rasing, T.; Shen, Y. R. Orientation of
Molecular Monolayers at the Liquid−Liquid Interface As Studied by
Optical Second Harmonic Generation. Langmuir 1988, 4, 452−454.
(23) Tamburello-Luca, A. A.; Hebert, P.; Brevet, P. F.; Girault, H. H.
Resonant-Surface Second-Harmonic Generation Studies of Phenol
Derivatives at Air/Water and Hexane/Water Interfaces. J. Chem. Soc.,
Faraday Trans. 1996, 92, 3079−3085.
(24) Salafsky, J. S.; Eisenthal, K. B. Second Harmonic Spectroscopy:
Detection and Orientation of Molecules at a Biomembrane Interface.
Chem. Phys. Lett. 2000, 319, 435−439.
(25) Gonella, G.; Dai, H.-L.; Fry, H. C.; Therien, M. J.; Krishnan, V.;
Tronin, A.; Blasie, J. K. Control of the Orientational Order and
Nonlinear Optical Response of the “Push−Pull” Chromophore RuPZn
via Specific Incorporation into Densely Packed Monolayer Ensembles
of an Amphiphilic 4-Helix Bundle Peptide: Second Harmonic
Generation at High Chromophore Densities. J. Am. Chem. Soc.
2010, 132, 9693−9700.
(26) Mitchell, S. A. Indole Adsorption to a Lipid Monolayer Studied
by Optical Second Harmonic Generation. J. Phys. Chem. B 2009, 113,
10693−10707.
(27) Simpson, G. J.; Rowlen, K. L. An SHG Magic Angle:
Dependence of Second Harmonic Generation Orientation Measurements on the Width of the Orientation Distribution. J. Am. Chem. Soc.
1999, 121, 2635−2636.
(28) Sen, S.; Yamaguchi, S.; Tahara, T. Different Molecules
Experience Different Polarities at the Air/Water Interface. Angew.
Chem., Int. Ed. 2009, 48, 6439−6442.
(29) Rao, Y.; Hong, S.-Y.; Turro, N. J.; Eisenthal, K. B. Molecular
Orientational Distribution at Interfaces Using Second Harmonic
Generation. J. Phys. Chem. C 2011, 115, 11678−11683.
(30) Kemnitz, K.; Bhattacharyya, K.; Hicks, J. M.; Pinto, G. R.;
Eisenthal, K. B.; Heinz, T. F. The Phase of Second-Harmonic Light
Generated at an Interface and Its Relation to Absolute Molecular
Orientation. Chem. Phys. Lett. 1986, 131, 285−290.
(31) Benjamin, I. Molecular Structure and Dynamics at Liquid−
Liquid Interfaces. Annu. Rev. Phys. Chem. 1997, 48, 407−451.
(32) Rivera, J.; McCabe, C.; Cummings, P. Molecular Simulations of
Liquid−Liquid Interfacial Properties: Water−n-Alkane and Water−
Methanol−n-Alkane Systems. Phys. Rev. E 2003, 67, No. 011603.
(33) Wang, Y.; Hodas, N. O.; Jung, Y.; Marcus, R. A. Microscopic
Structure and Dynamics of Air/Water Interface by Computer
SimulationsComparison with Sum-Frequency Generation Experiments. Phys. Chem. Chem. Phys. 2011, 13, 5388−5393.
(34) Wernersson, E.; Heyda, J.; Vazdar, M.; Lund, M.; Mason, P. E.;
Jungwirth, P. Orientational Dependence of the Affinity of
Guanidinium Ions to the Water Surface. J. Phys. Chem. B 2011, 115,
12521−12526.
(35) Razavi, S.; Koplik, J.; Kretzschmar, I. Molecular Dynamics
Simulations: Insight into Molecular Phenomena at Interfaces.
Langmuir 2014, 30, 11272−11283.
(36) Pohorille, A.; Benjamin, I. Molecular Dynamics of Phenol at the
Liquid−Vapor Interface of Water. J. Chem. Phys. 1991, 94, 5599−5605.
(37) Kundu, A.; Watanabe, H.; Yamaguchi, S.; Tahara, T. Agreement
between Experimentally and Theoretically Estimated Orientational
Distributions of Solutes at the Air/Water Interface. J. Phys. Chem. C
2013, 117, 8887−8891.
(38) Zheng, H.-d.; Wu, F.-d.; Wang, B.-y.; Wu, Y.-x. Molecular
Dynamics Simulation on the Interfacial Features of Phenol Extraction
by TBP/Dodecane in Water. Comput. Theor. Chem. 2011, 970, 66−72.
(39) Beierlein, F.; Krause, A.; Jaeger, C. M.; Fita, P.; Vauthey, E.;
Clark, T. Molecular-Dynamics Simulations of Liquid Phase Interfaces:
Understanding the Structure of the Glycerol/Water−Dodecane
System. Langmuir 2013, 29, 11898−11907.
(40) Richert, S.; Mosquera Vazquez, S.; Grzybowski, M.; Gryko, D.
T.; Kyrychenko, A.; Vauthey, E. Excited-State Dynamics of an
Environment-Sensitive Push−Pull Diketopyrrolopyrrole: Major Differ-
for providing the compounds for study. This work was
supported by the Swiss National Science Foundation through
Project 200020-147098 and the University of Geneva. D.S. and
D.K. also thank the Swiss Federal Commission for Scholarships
and the Ministry of Education and Science of Ukraine,
respectively, for supporting their stays in Switzerland.
■
REFERENCES
(1) Volkov, A. G. E. Liquid Interfaces in Chemical, Biological, and
Pharmaceutical Applications; Marcel Dekker: New York, 2009.
(2) Watarai, H.; Teramae, N.; Sawada, T. Interfacial Nanochemistry;
Kluwer Academic: New York, 2005.
(3) Adamson, A. W.; Gast, A. P. Physical Chemistry of Surfaces, 6th
ed.; Wiley: New York, 1997.
(4) Breslow, R. Hydrophobic Effects on Simple Organic Reactions in
Water. Acc. Chem. Res. 1991, 24, 159−164.
(5) Narayan, S.; Muldoon, J.; Finn, M. G.; Fokin, V. V.; Kolb, H. C.;
Sharpless, K. B. “On Water”: Unique Reactivity of Organic
Compounds in Aqueous Suspension. Angew. Chem., Int. Ed. 2005,
44, 3275−3279.
(6) Jung, Y.; Marcus, R. A. On the Theory of Organic Catalysis “on
Water”. J. Am. Chem. Soc. 2007, 129, 5492−5502.
(7) Corn, R. M.; Higgins, D. A. Optical Second Harmonic
Generation as a Probe of Surface Chemistry. Chem. Rev. 1994, 94,
107−125.
(8) Eisenthal, K. B. Liquid Interfaces Probed by Second-Harmonic
and Sum-Frequency Spectroscopy. Chem. Rev. 1996, 96, 1343−1360.
(9) Richert, S.; Fedoseeva, M.; Vauthey, E. Ultrafast Photoinduced
Dynamics at Air/Liquid and Liquid/Liquid Interfaces. J. Phys. Chem.
Lett. 2012, 3, 1635−1642.
(10) Nihonyanagi, S.; Mondal, J. A.; Yamaguchi, S.; Tahara, T.
Structure and Dynamics of Interfacial Water Studied by HeterodyneDetected Vibrational Sum-Frequency Generation. Annu. Rev. Phys.
Chem. 2014, 64, 579−603.
(11) Richmond, G. L. Molecular Bonding and Interactions at
Aqueous Surfaces as Probed by Vibrational Sum Frequency Spectroscopy. Chem. Rev. 2002, 102, 2693−2724.
(12) Shultz, M. J.; Schnitzer, C.; Simonelli, D.; Baldelli, S. Sum
Frequency Generation Spectroscopy of the Aqueous Interface. Ionic
and Soluble Molecular Solutions. Int. Rev. Phys. Chem. 2000, 19, 123−
153.
(13) McGuire, J.; Shen, Y. R. Ultrafast Vibrational Dynamics at Water
Interfaces. Science 2006, 313, 1945−1948.
(14) Zhang, Z.; Piatkowski, L.; Bakker, H. J.; Bonn, M. Ultrafast
Vibrational Energy Transfer at the Water/Air Interface Revealed by
Two-Dimensional Surface Vibrational Spectroscopy. Nat. Chem. 2011,
3, 888−893.
(15) Steel, W. H.; Walker, R. A. Measuring Dipolar Width across
Liquid−Liquid Interfaces with “Molecular Rulers”. Nature 2003, 424,
296−299.
(16) Fujiwara, K.; Wada, S.; Monjushiro, H.; Watarai, H. IonAssociation Aggregation of an Anionic Porphyrin at the Liquid/Liquid
Interface Studied by Second Harmonic Generation Spectroscopy.
Langmuir 2006, 22, 2482−2486.
(17) McArthur, E. A.; Eisenthal, K. B. Ultrafast Excited-State Electron
Transfer at an Organic Liquid/Aqueous Interface. J. Am. Chem. Soc.
2006, 128, 1068−1069.
(18) Onorato, R. M.; Otten, D. E.; Saykally, R. J. Adsorption of
Thiocyanate Ions to the Dodecanol/Water Interface Characterized by
UV Second Harmonic Generation. Proc. Natl. Acad. Sci. U.S.A. 2009,
106, 15176−15180.
(19) Fedoseeva, M.; Richert, S.; Vauthey, E. Excited-State Dynamics
of Organic Dyes at Liquid−Liquid Interfaces. Langmuir 2012, 28,
11291−11301.
(20) Yamaguchi, S.; Tahara, T. Precise Electronic χ(2) Spectra of
Molecules Adsorbed at an Interface Measured by Multiplex Sum
Frequency Generation. J. Phys. Chem. B 2004, 108, 19079−19082.
13875
dx.doi.org/10.1021/la503121g | Langmuir 2014, 30, 13869−13876
Langmuir
Article
ences between the Bulk Solution Phase and the Dodecane/Water
Interface. J. Phys. Chem. B 2014, 118, 9952−9963.
(41) Fedoseeva, M.; Letrun, R.; Vauthey, E. Excited-State Dynamics
of Rhodamine 6G in Aqueous Solution and at the Dodecane/Water
Interface. J. Phys. Chem. B 2014, 118, 5184−5193.
(42) Fedoseeva, M.; Fita, P.; Vauthey, E. Excited-State Dynamics of
Charged Dyes at Alkane/Water Interfaces in the Presence of Salts and
Ionic Surfactants. Langmuir 2013, 29, 14865−14872.
(43) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A., et al. Gaussian 09, revision C1; Gaussian, Inc.:
Wallingford, CT, 2009.
(44) Pronk, S.; Páll, S.; Schulz, R.; Larsson, P.; Bjelkmar, P.;
Apostolov, R.; Shirts, M. R.; Smith, J. C.; Kasson, P. M.; van der Spoel,
D.; et al. GROMACS 4.5: A High-Throughput and Highly Parallel
Open Source Molecular Simulation Toolkit. Bioinformatics 2013, 29,
845−854.
(45) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R.
W.; Klein, M. L. Comparison of Simple Potential Functions for
Simulating Liquid Water. J. Chem. Phys. 1983, 79, 926−935.
(46) Schüttelkopf, A. W.; van Aalten, D. M. F. PRODRG: A Tool for
High-Throughput Crystallography of Protein−Ligand Complexes.
Acta Crystallogr., Sect. D 2004, 60, 1355−1363.
(47) Hess, B. P-Lincs: A Parallel Linear Constraint Solver for
Molecular Simulation. J. Chem. Theory Comput. 2008, 4, 116−122.
(48) Essmann, U.; Perera, L.; Berkowitz, M. L.; Darden, T.; Lee, H.;
Pedersen, L. G. A Smooth Particle Mesh Ewald Method. J. Chem. Phys.
1995, 103, 8577−8593.
(49) Bussi, G.; Zykova-Timan, T.; Parrinello, M. Isothermal−Isobaric
Molecular Dynamics Using Stochastic Velocity Rescaling. J. Chem.
Phys. 2009, 130, No. 074101.
(50) Berendsen, H. J. C.; Postma, J. P. M.; van Gunsteren, W. F.;
DiNola, A.; Haak, J. R. Molecular Dynamics with Coupling to an
External Bath. J. Chem. Phys. 1984, 81, 3684−3690.
(51) Humphrey, W.; Dalke, A.; Schulten, K. VMD: Visual Molecular
Dynamics. J. Mol. Graphics 1996, 14, 33−38.
(52) Ponomarev, O. A.; Brusil’tsev, Y. N.; Kotelevskii, S. I.; Mitina, V.
G.; Pedash, Y. F. Nature of the Excited States of Dialkylamino
Derivatives of Aromatic and Heteroaromatic Compounds with an
Annelated Oxazole Ring. Theor. Exp. Chem. 1991, 26, 605−611.
(53) Pant, D.; Le Guennec, M.; Illien, B.; Girault, H. H. The pH
Dependent Adsorption of Coumarin 343 at the Water/Dichloroethane
Interface. Phys. Chem. Chem. Phys. 2004, 6, 3140−3146.
(54) Benderskii, A. V.; Eisenthal, K. B. Aqueous Solvation Dynamics
at the Anionic Surfactant Air/Water Interface. J. Phys. Chem. B 2001,
105, 6698−6703.
(55) Zimdars, D.; Eisenthal, K. B. Static and Dynamic Solvation at
the Air/Water Interface. J. Phys. Chem. B 2001, 105, 3993−4002.
(56) Makarov, N. S.; Drobizhev, M.; Rebane, A. Two-Photon
Absorption Standards in the 550−1600 nm Excitation Wavelength
Range. Opt. Express 2008, 16, 4029−4047.
(57) Shen, Y. R. Surface Contribution versus Bulk Contribution in
Surface Nonlinear Optical Spectroscopy. Appl. Phys. B: Lasers Opt.
1999, 68, 295−300.
(58) Fita, P.; Fedoseeva, M.; Vauthey, E. Hydrogen-Bond-Assisted
Excited-State Deactivation at Liquid/Water Interfaces. Langmuir 2011,
27, 4645−4652.
(59) Brevet, P.-F. Surface Second Harmonic Generation; Presses
Polytechniques et Universitaires Romandes: Lausanne, Switzerland,
1997.
(60) Brand, C.; Meerts, W. L.; Schmitt, M. How and Why Do
Transition Dipole Moment Orientations Depend on Conformer
Structure? J. Phys. Chem. A 2011, 115, 9612−9619.
13876
dx.doi.org/10.1021/la503121g | Langmuir 2014, 30, 13869−13876